Computational Model for Urban Growth Using Socioeconomic Latent Parameters

TL;DR

This paper introduces a novel Hidden Markov Model-based approach for urban land use change prediction that incorporates socioeconomic factors, demonstrating improved accuracy over traditional Markov Chain models in Pune, India.

Contribution

The paper presents the first application of Hidden Markov Models to land use change modeling, integrating socioeconomic variables with spatio-temporal data for enhanced urban growth prediction.

Findings

HMM-based model outperforms Markov Chain in prediction accuracy

Socioeconomic factors significantly influence land use changes

Integrated models improve urban growth visualization

Abstract

Land use land cover changes (LULCC) are generally modeled using multi-scale spatio-temporal variables. Recently, Markov Chain (MC) has been used to model LULCC. However, the model is derived from the proportion of LULCC observed over a given period and it does not account for temporal factors such as macro-economic, socio-economic, etc. In this paper, we present a richer model based on Hidden Markov Model (HMM), grounded in the common knowledge that economic, social and LULCC processes are tightly coupled. We propose a HMM where LULCC classes represent hidden states and temporal fac-tors represent emissions that are conditioned on the hidden states. To our knowledge, HMM has not been used in LULCC models in the past. We further demonstrate its integration with other spatio-temporal models such as Logistic Regression. The integrated model is applied on the LULCC data of Pune district in the state of Maharashtra (India) to predict and visualize urban LULCC over the past 14 years. We observe that the HMM integrated model has improved prediction accuracy as compared to the corresponding MC integrated model